Scores are global-scale research estimates. R marks work in progress; none are travel advice.
Found a hotel? Check the area around it before booking.
Search sends your query and selected city to OpenStreetMap Nominatim. Don't enter private addresses or sensitive information. Search policy
Know before you roam
Compare Berlin areas using official recorded-theft evidence, with broader travel-safety context and source details.
Safrom's citywide research estimate is 83/100. Higher scores mean safer on the model's scale. Evidence as of . Confidence: C.
143 areas with estimates. Area evidence as of .
recorded_thefts_per_1000_residents · 2025-01-01 / 2025-12-31.
Official register population denominator differs from the federal census used in the headline PKS rate. Per-resident theft rates are influenced by visitors and commuters. Unlocated incidents remain in city totals and are not assigned to individual areas. The source city total exceeds the sum of published area and unknown rows by one theft; no area value is changed. Theft is a specific property-risk input; no time-of-day or other local risk dimension is inferred. The city score uses national priors. Local theft rates adjust only the property dimension across 143 Bezirksregionen. Derived local dataset is CC BY-SA; retain Polizei Berlin attribution and share-alike. LOR boundaries have their separate CC BY 3.0 attribution.
Download the scored safety layer and calculation evidence (GeoJSON)
Official register population denominator differs from the federal census used in the headline PKS rate. Per-resident theft rates are influenced by visitors and commuters. Unlocated incidents remain in city totals and are not assigned to individual areas. The source city total exceeds the sum of published area and unknown rows by one theft; no area value is changed. Theft is a specific property-risk input; no time-of-day or other local risk dimension is inferred. The city score uses national priors. Local theft rates adjust only the property dimension across 143 Bezirksregionen. Derived local dataset is CC BY-SA; retain Polizei Berlin attribution and share-alike. LOR boundaries have their separate CC BY 3.0 attribution.
Boundaries: Amt für Statistik Berlin-Brandenburg · Berlin LOR Bezirksregionen, 2023 (CC BY 3.0 (boundaries); CC BY-SA (derived theft dataset)).
Official register population denominator differs from the federal census used in the headline PKS rate. Per-resident theft rates are influenced by visitors and commuters. Unlocated incidents remain in city totals and are not assigned to individual areas. The source city total exceeds the sum of published area and unknown rows by one theft; no area value is changed. Theft is a specific property-risk input; no time-of-day or other local risk dimension is inferred. The city score uses national priors. Local theft rates adjust only the property dimension across 143 Bezirksregionen. Derived local dataset is CC BY-SA; retain Polizei Berlin attribution and share-alike. LOR boundaries have their separate CC BY 3.0 attribution.
Natural=4; current conflict intensity=1.1; infrastructure=0.7; access to health care=0.6. Each published 0–10 component is multiplied by ten for the existing research risk scale.; Country humanitarian-risk components are weak priors, not measured Berlin street or individual tourist risks.; European Union, CC BY 4.0; component extraction and research transformation by Safrom.
Homicide: 0.9 per 100,000 (2023); theft: 1,241.4 per 100,000 (2022). Country observations, not Berlin rates.; Property prior uses the same 2022 UN theft series against France (2,011.4), anchored at the existing France research risk 45: 45 + bounded(10 × log2(rate/reference), ±20). This is an explicit weak model calibration, not an official score.; Recording definitions and reporting propensity differ between countries. This country comparison does not create locally observed risks; geographic-fit confidence is 0.2.
34 deaths per million residents, equivalent to 3.4 per 100,000. Preliminary full-year estimate published 2026-03-24; not Berlin traffic exposure.; European Union, CC BY 4.0; unit conversion and research transformation by Safrom.
Official register population denominator differs from the federal census used in the headline PKS rate.; Per-resident theft rates are influenced by visitors and commuters.; Unlocated incidents remain in city totals and are not assigned to individual areas. The source city total exceeds the sum of published area and unknown rows by one theft; no area value is changed.; Theft is a specific property-risk input; no time-of-day or other local risk dimension is inferred.; CC BY-SA; Polizei Berlin. Local aggregation and transformation by Safrom; redistribution of the derived dataset must preserve attribution and share-alike.